Re: [PATCH] mount.nfs: Teach mount.nfs about the mountversion option

[Steve Dickson <SteveD@redhat.com>]

On 07/06/2015 06:29 PM, Trond Myklebust wrote:
> On Mon, Jul 6, 2015 at 6:21 PM, Kinglong Mee <kinglongmee@gmail.com> wrote:
>> On 7/6/2015 23:19, Trond Myklebust wrote:
>>> On Mon, Jul 6, 2015 at 10:42 AM, Steve Dickson <steved@redhat.com> wrote:
>>>> When the mountversion option is used, there should
>>>> not be any mount negotiations with the server.
>>>>
>>>> Also, when the option is used, its know that the mount
>>>> is a v4 mount and a V_SPECFIC type.
>>>>
>>>> Signed-off-by: Steve Dickson <steved@redhat.com>
>>>> ---
>>>>  utils/mount/network.c | 7 ++++++-
>>>>  1 file changed, 6 insertions(+), 1 deletion(-)
>>>>
>>>> diff --git a/utils/mount/network.c b/utils/mount/network.c
>>>> index b5ed850..ebc39d3 100644
>>>> --- a/utils/mount/network.c
>>>> +++ b/utils/mount/network.c
>>>> @@ -92,6 +92,7 @@ static const char *nfs_version_opttbl[] = {
>>>>         "v4",
>>>>         "vers",
>>>>         "nfsvers",
>>>> +       "minorversion",
>>>>         NULL,
>>>>  };
>>>>
>>>> @@ -1272,7 +1273,11 @@ nfs_nfs_version(struct mount_options *options, struct nfs_version *version)
>>>>         if (!(version->major = strtol(version_val, &cptr, 10)))
>>>>                 goto ret_error;
>>>>
>>>> -       if (version->major < 4)
>>>> +       if (strcmp(nfs_version_opttbl[i], "minorversion") == 0) {
>>>> +               version->v_mode = V_SPECIFIC;
>>>> +               version->minor = version->major;
>>>> +               version->major = 4;
>>>> +       } else if (version->major < 4)
>>>>                 version->v_mode = V_SPECIFIC;
>>>>
>>>>         if (*cptr == '.') {
>>>>
>>>
>>> Doesn't this cause a dependency on the ordering of 'nfsvers'/'vers'
>>> and 'minorversion' in your mount options string?
>>
>> Without this patch, there is a dependency on the ordering now.
>> mount with "-o minorversion=1", nfs gets strings as "minorversion=1,vers=4.2"
>>
>> If nfs-utils doesn't parse minorversion, the order dependency also exist.
>>
>> I agree with parsing minorversion in nfs-utils.
>>
> 
> I was rather hoping we could get rid of 'minorversion' soon. It is a
> wart, and we've already agreed to deprecate it.
I agree with the deprecation, but I think that is a different issue...

steved.